Abstract: A neural network operation apparatus and method implementing quantization is disclosed. The neural network operation method may include receiving a weight of a neural network, a candidate set of quantization points, and a bitwidth for representing the weight, extracting a subset of quantization points from the candidate set of quantization points based on the bitwidth, calculating a quantization loss based on the weight of the neural network and the subset of quantization points, and generating a target subset of quantization points based on the quantization loss.

Abstract: The present invention relates to a benzamide derivative, a preparation method therefor, and a pharmaceutical composition comprising same as an active ingredient for prevention or treatment of cancer. The benzamide derivative provided in an aspect of the present invention can be used for preventing or treating cancer by suppressing EGFR mutation, and exhibits a remarkable synergy effect on anticancer activity when administered in combination with an EGFR antagonist such as Cetuximab, thus finding advantageous uses as an anticancer agent.

Abstract: Disclosed are a direct-current transmission line protection method and system based on pure current characteristics.

Abstract: The disclosure provides a panel of biomarkers that individually or in combination can indicate the presence of sepsis as distinguishable from other non-infection related inflammatory conditions. The disclosed biomarkers and related reagents and kits provide strategies for detecting, treating, and monitoring sepsis in subjects. In aspect, the disclosure provides a method for detecting sepsis, comprising contacting a biological sample obtained from the subject with an affinity reagent that specifically binds to one or more of the disclosed novel biomarkers, and detecting differential expression of the one or more biomarkers by detecting binding of the affinity reagent to the biomarker. The method can incorporate use of additional known biomarkers. The method can further comprise treating a subject determined to have sepsis. In some embodiments, the subject is a human subject less than 20 years old.

Abstract: A method of blind spot detection is provided. The method of blind spot detection is used for a vehicle including a first vehicle body and a second vehicle body dragged by the first vehicle body. The method of blind spot detection comprises the following steps: obtaining a turning angle information of the second vehicle body relative to the first vehicle body by a first sensor while the vehicle is moving; determining a predetermined information related to the vehicle and/or a second sensor; dynamically defining a blind spot area around the vehicle according to the turning angle information and the predetermined information; and receiving a sensing signal regarding an object around the vehicle from the second sensor to determine whether the object is located in the blind spot area.

Abstract: Compositions and methods related to the synthesis and application of poly(aryl ether)s are generally described.

Abstract: A prediction device extracts each predictive feature vector expressing a feature from a peptide that is a target for biostability prediction. The prediction device generates a predicted value of biostability of the prediction target cyclic peptide by inputting plural predictive feature vectors into a trained model pre-trained to output a predicted value of peptide biostability.

Abstract: The present disclosure provides for base editors which satisfy a need in the art for installation of targeted transversions of guanine (G) to thymine (T), or correspondingly, transversions of adenine (A) to cytosine (C). The domains of the disclosed base editors include a nucleic acid programmable DNA binding protein and a guanine oxidase or a guanine methyltransferase. The base editors may be engineered through the use of continuous or non-continuous evolution systems. In particular, the present disclosure provides for guanine-to-thymine (or cytosine-to-adenine) base editors that can install single-base trans version mutations. In addition, methods for targeted nucleic acid editing are provided. Further provided are pharmaceutical compositions comprising, and vectors and kits useful for the generation of, guanine-to-thymine base editors. Cells containing such vectors and cells containing base editors and guide RNAs are also provided.

Abstract: Provided herein are genetic circuits and cell state classifiers for detecting the microRNA profile of a cell. The cell state classifiers of the present disclosure are designed to incorporate multiple genetic circuits integrated together by transcriptional or translational control. Multiple inputs can be sensed simultaneously by coupling their detection to different portions of the genetic circuit such that the output molecule is produced only when the correct input profile of miRNAs is detected. The genetic circuits and cell state classifiers may be used in various applications (e.g., therapeutic or diagnostic applications).

Abstract: The present invention improves a heat dissipation property of a semiconductor device by transferring hexagonal boron nitride (hBN) with a two-dimensional nanostructure to the semiconductor device. A semiconductor device of the present invention includes a substrate having a first surface and a second surface, a semiconductor layer formed on the first surface of the substrate, an hBN layer formed on at least one surface of the first surface and the second surface of the substrate, and a heat sink positioned on the second surface of the substrate. A radiation rate of heat generated during driving of an element is increased to decrease a reduction in lifetime of a semiconductor device due to a temperature increase. The semiconductor device has a structure and configuration which are very effective in improving a rapid temperature increase due to heat generated by high-power semiconductor devices.

Abstract: An autonomous vehicle includes first and second road surface image obtaining devices that are located on a bottom surface of the vehicle body and obtain images of a road surface below the vehicle body, respectively. The autonomous vehicle also includes a memory unit that stores a map image of the road surface, the map image being associated with geographical location information. The autonomous vehicle further includes first and second self-location estimating units that each compare a feature extracted from the image of the road surface obtained by corresponding one of the first and second road surface image obtaining devices with a feature extracted from the map image, thereby estimating a self-location of the autonomous vehicle.

Abstract: A transistor includes a substrate; a pair of constant current forming regions provided in the substrate; a pair of source/drain regions respectively provided on the pair of constant current forming regions in the substrate; and a gate structure provided between the pair of source/drain regions, wherein any one of the constant current forming regions immediately adjacent to any one of the pair of source/drain regions serving as a drain forms a constant current between the any one of the pair of source/drain region serving as the drain and the any one of the constant current forming regions.

Abstract: Disclosed herein are a method for making a montage based on a dialogue and an apparatus using the same. The method includes extracting a feature for making a montage based on the dialogue between at least one inquirer and at least one respondent, generating a facial feature image based on the feature, and making a montage based on the facial feature image.

Abstract: A system and a method for training a semantic segmentation model includes obtaining a plurality of sets of images each having an index z for visibility, iteratively training the model. Iteratively training the model includes (a) for each z above 1, obtaining preliminary semantic segmentation labels for each image of the set of images of index z?1 by applying the model to each image of the set of images of index z?1, (b) processing each preliminary semantic segmentation labels using semantic segmentation labels obtained using the model on a selected image of index 1, and obtaining processed semantic segmentation labels, (c) training the model using the set of images of index z?1 and the associated processed semantic segmentation labels, and (d) performing steps (a) to (c) for z+1.

Abstract: Provided is a transistor including: a constant current formation layer; a channel layer provided on the constant current formation layer; a pair of source/drain regions spaced apart from each other, with the channel layer therebetween on the constant current formation layer; a gate electrode provided on the channel layer; and a gate ferroelectric film provided between the gate electrode and the channel layer.

Abstract: Disclosed are bispecific compounds (degraders) that target CDK7 for degradation. Also disclosed are pharmaceutical compositions containing the degraders and methods of using the compounds to treat disease.

Abstract: Disclosed is a preparation method for preparing a measurement sample comprising an aggregate of metal nanoparticles having an analyte bound thereto, the preparation method comprising: contacting the analyte with a linker to bind the analyte to the linker; and contacting the linker that has been bound to the analyte with the metal nanoparticles to bind the linker to the metal nanoparticles.

Abstract: According to one embodiment, a magnetic sensor includes a first sensor part. The first sensor part includes a first magnetic member, a first counter magnetic member, and a first magnetic element including one or a plurality of first extension parts. The first extension part includes a first magnetic layer, a first counter magnetic layer, and a first nonmagnetic layer. The first magnetic layer includes a first portion, a first counter portion, and a first middle portion. The first middle portion is between the first portion and the first counter portion. The first nonmagnetic layer is between the first counter magnetic layer and at least a portion of the first middle portion. An electrical resistance of the first magnetic element has first to third values when first to third magnetic fields are applied to the first magnetic element. The first value is greater than the second and third values.

Abstract: An encoding method includes receiving input data represented by a 16-bit half floating point, adjusting a number of bits of an exponent and a mantissa of the input data to split the input data into 4-bit units, and encoding the input data in which the number of bits has been adjusted such that the exponent is a multiple of “4”.

Abstract: A transistor includes: a substrate; a constant current formation layer provided on the substrate; a pair of source/drain patterns provided on the constant current formation layer; a gate electrode provided between the pair of source/drain patterns; a channel pattern extending in a direction between the pair of source/drain patterns; and a gate insulating layer surrounding the channel pattern, wherein the channel pattern penetrates the gate insulating layer and the gate electrode and is electrically connected to the source pattern and the drain pattern, the gate insulating layer separates the channel pattern and the gate electrode from each other, the constant current formation layer generates a constant current between the drain pattern and the substrate, and the constant current is independent from a gate voltage applied to the gate electrode.